Operational implications

Providing technical support
In order to maximise a demand oriented community based approach, communities need a variety of types of support; information, guidance, organisational support and capacity building. Technical support to communities should include items such as the organisation of community meetings, allowing communities to make their own decisions, deciding what they want to do about them, and explaining how proposed projects can be of help. The role for support agencies is one of explaining the range of technical options, the implications for each option, financial and other costs, and how decisions may affect the wider environment.

In many cases, government agencies may not be best placed to provide the type of support needed, or at the time it is needed, and these skills may be found in other intermediaries, most probably in NGO's or consultants. A key challenge for SSA is the need to understand what criteria can be used to determine if a particular intermediary is suitable for government and donors in providing technical support (Wright, 1997).

The shift in thinking away from the State as the sole provider of services and goods towards a role as enabling and facilitating other agencies to provide services and support is well documented. In this arrangement, certain organisations act as intermediaries or brokers between communities and local governments, often getting the process of development started, initiating dialogue between communities and local authorities and translating government policy to communities. In many cities, NGO's have successfully involved themselves in people's development processes and developed an expertise in this field (UNCHS/CityNet, 1997). Within the sphere of infrastructure provision, Choguill (1997), for example, cites local and international NGO's playing a key role in assisting communities develop their basic services as one of the ten steps to sustainable urban infrastructure.

There are many examples from the literature of NGO's acting in an intermediary role, providing the types of technical support described in SSA. In the Gajapura Bosevana case study in Sri Lanka, (Hewawasam et al, 1997:7) describes the role that SEVANATHA took in enabling community mobilisation, '...SEVANATHA was responsible for community mobilisation, planning and organising the community education and training programmes, finding funds for construction, providing technical advice for the community during construction work, and training..of community leaders on maintenance of community infrastructure.' Wright (1997) cites two examples to further illustrate this point, the well documented Orangi Pilot Project in Karachi Pakistan (where OPP has built capacity to the extent that local organisations are able to plan and finance their own latrines and house drains; installed in-house sanitary latrines, house drains and shallow sewers), and the work of the Dominican Institute of Integral Development (IDDI) working in La Zurza settlement (which led to creation of a local CBO, SODIZUR, and a three stage programme to improve basic services and infrastructure). These examples demonstrate NGO's roles as intermediaries in the development process.

Of greater interest for the application of SSA are lessons which are learnt from this role. The impact on NGO's of their increasingly prominent role in the development process is explored by Bebbington & Mitlin (1996). Their research indicated NGO capacity building should be seen not only as a means or an end, but as a process. The implications of this for individual NGO’s is a greater emphasis on a coherent institutional structure and operation. As NGO’s adapt to their 'gap filling' role in taking over former State based activities, institutional coherence becomes critical since the demands of this role require new operational skills, and make it harder for the NGO structures and activities to continue to match its original mission statement.

In the Gomti River Pollution Control project at Lucknow, a multi-consultancy approach (including NGO's) was adopted for the delivery of project inputs. Although well intentioned, this strategy led to difficulties in that a hierarchy of partners was established in which NGO's did not feature prominently (Harvey, 1996). Two key lessons to emerge were to treat intermediaries such as NGO's and consultants as separate agencies, rather than similar organisation types with similar modes of operation, and to consider how 'guardian' consultants could help develop NGO inputs.

Some commentators argue against the received wisdom of NGO's as the natural partner to facilitate intermediary functions (Varley et al, 1996). While it is true that NGO's are seen as being more responsive to locally expressed demand and more effective in encouraging participation, there are concerns arising from any support role they might provide. These include:

In the same way that much of the existing literature recommends new working practices in local government, Varley et al argue that NGO's also need to be pragmatic in their partnerships, being open to working with government and avoiding a confrontational atmosphere that may discourage rather than enhance their participation in local decisions/programmes. In each particular case, NGO's should be included in the institutional analysis for the programme, and their relative merits and disadvantages assessed. Procedures for integrating technical support from other agencies need to be flexible and there needs to be commitment amongst donors and government agencies to work with NGO's. Many NGO's are sceptical about government while at the same time many government officials are suspicious of NGO's (UNCHS/CityNet, 1997). A trust building process is required in these collaborations. If this is absent, then the NGO may be forced into a service delivery rather than capacity building role. In this case, project priorities, and timing of implementation and procurement procedures may need to be changed to provide greater flexibility to allow NGO's to design and implement their own programmes.

Carrol et al (1995) provide a useful overview of the key criteria used to help determine whether a particular NGO is likely to be a suitable intermediary for governments and donors. Several indicators that an NGO meet the preconditions involved in effective partnerships include: flat management structures with decentralised authority; organisational structures at community level for decision making; use of iterative planning with local community; contributions of cash, labour or materials by local community; staff recruitment criteria, incentives and training that supports participation; positive perceptions of the NGO by local community leaders; turnover of client groups as they 'graduate' over time.

However, in some cases, the qualities that make an NGO suitable as an intermediary may clash with government or donor requirements. In Zambia, the Squatter Upgrading Project agreed in principle to pursue long term community development goals by promoting active beneficiary participation. However, a key precondition was that if the collective self help approach was used by two intermediary NGO's and interfered with the project schedule, then contractors would be employed to complete the work rather than the NGO.

Key points:

Widening technological options
One of the lessons of the last two decades is that a lower cost technology is not in itself a guarantor of sustainable investments in sanitation services. Strategic sanitation's emphasis on demand requires consideration of not only lower cost technologies but a wider choice of technological options across the full cost range than was generally the case in the past.

A framework for analysing technological options is proposed which is based on a three way division of sanitation infrastructure levels: in-house, feeder and trunk. Levels of demand need to be assessed for each division (Wright, 1997).

A distinctive feature of demand oriented programmes is that users are allowed to make choices from a range (or 'menu') of sanitation options with appropriate price tags. However, there are few practical examples from the literature which exemplify this 'menu' approach to sanitation technologies. In many notable cases, including OPP, Bangladesh SIP, the Gajabapur-Bosevana settlement in Sri Lanka, and the Cuttack Urban Services Improvement Project it is clear that individual communities were not made aware of a range of technological options, their differing costs and suitability to specific areas. In effect, choices were made on behalf of the user community by sector professionals.

The idea of widening technological options as proposed through SSA raises a number of difficulties which can be addressed from two perspectives: the users' and service providers.

Ideally, choice is based on an assessment of different options which is informed by timely and relevant information. It also assumes that the person making the choice understands the consequences (technical, managerial, financial) of that option. In practice, this is rarely the case, and choices are made on incomplete, or poorly informed knowledge. There is little discussion of this point when advocating the widening of technological options.

On one level, 'choice' for users may be limited because of the inadequacy of information about alternative options available at the local level. In this context, users may express a preference for systems which they see around them and which are familiar. However, this is not to be confused with idea of choice based on a reasoned assessment of known alternatives. Furthermore, SSA fails to recognise the potential problem of partially informed user choices. Where community knowledge is based on an incomplete or inadequate understanding of the technical consequences of different systems in particular circumstances (i.e., choice of sewerage system in area without piped water supply), then there is a danger that user choices will affect system sustainability.

In contrast to these points, the experiences of NSDF/SPARC/MM in various Indian cities suggest otherwise, and Kurup (1997) argues that these programmes indicate that communities are capable of making technical decisions about (and between) sanitation systems. Analysis of lessons learnt indicate the poor are capable of:

What is required is that the debate about technological options should not be overly rigid (Tayler, 1996). It is apparent that some community choices may have limitations which are obvious and serious; in other cases, it may be that the preferred options of both professionals and the people prove to have weaknesses. In these instances, a degree of negotiation takes place, "...in which the sharing of views and information results in a third, more appropriate option being found " (Tayler, 1996:12).

From the service providers perspective, one of the assumptions underpinning widening technological options is that engineers and other professionals are willing and able to adopt a wider range of sanitation options as 'appropriate'. In practice, this is not always the case. Didactic methods of education and university courses based on the problems of western industrialised countries tend to constrain innovative engineering approaches. The prevailing work culture within government or municipal sanitation departments acts against the adoption of 'alternative' engineering designs. Although most government service positions bring with them a high degree of employee security, many officers are unwilling to take professional risks and well established procedures are adhered to scrupulously (Evans, 1995). Although standard technology designs and use of technical manuals help to reduce design costs, they also stifle innovation. Standard designs are advocated not because they are necessarily the most appropriate solution, but because in the event of failure it is possible to shift blame onto the wider 'system'.

This point is reinforced in research examining the performance of on-plot sanitation in low income urban areas, where it is observed that, "...[there is] an underlying feeling amongst some authorities and sector professionals that whilst on-plot sanitation was appropriate for rural areas, it was generally unsuitable in the urban context, unless viewed as a (preferably short term) route to 'better' forms of sanitation. All too often, assessments and judgements on the effectiveness and appropriateness [of plot systems] are made from a technologically biased and purely external perspective" (Cotton and Saywell, 1998:3).

Irrespective of the user and service provider perspectives, in practice, choice is often constrained by a variety of factors, including technical, social, institutional and other issues (UNDP-World Bank Water and Sanitation Programme, 1996). Technical problems vary according to the system in question, but for sewerage may relate to flat topography and the lack of necessary gradients. Cotton (1998b) reinforces this point with lessons drawn from Cuttack, where user choice was limited because the topography of the city restricted horizontal unbundling of technologies. Other concerns include the integrity of materials and components used in sewered options, or the design standards for sewer diameters, gradients, and the siting of manholes and chambers. On plot systems present different technical issues, relating mainly to plot size, disposal of sullage water, impact of groundwater pollution from densely spaced pit latrine, operation of twin and double pit systems and desludging and disposal of latrines. Many of these issues are discussed in Cotton and Saywell (1998).

Institutional factors can limit choices when overly complex systems are advocated, and the technical capacity for O&M and management is lacking or cannot be sustained in the future. Cotton (1997) develops this point and questions the validity of conventional engineering planning and design which focuses on technical options and their technical feasibility as the primary determinants of success in implementation and sustainability. The requirements of a particular technical option ultimately need to fit with the local capacity that exists both within local government and local communities to cope with the infrastructure. Hence, the key factor in widening technological options needs to be a consideration of the technical, financial and managerial capacity to cope.

Key points:

Key questions:

Assessing sanitation demand
A demand oriented approach to sanitation is responsive, rather than prescriptive, one in which stakeholders are drawn into the process of decision making at all stages, including assessments of sanitation demand. Instead of governments or service utilities deciding which peri-urban communities should be provided with what type and level of service, the decisions are made jointly, through consultation and negotiations among all interested beneficiaries. All participants in a sanitation programme need to be aware of the implications of the options open to them, and it is the role of the implementing agency to explain technology options, costs, financing packages, and institutional arrangements involved in these options. Implementing agencies may set demand based eligibility criteria for sanitation interventions (i.e., community contributions to project costs, or O&M activities); alternatively, needs based criteria (i.e. health and poverty indicators) may be used to prioritise regions, but only for those who have expressed a willingness to pay (Wright, 1997).

Misjudgements about consumer demand are important contributing factors to poor sanitation system design and performance (World Bank, 1992). Demand assessment is critical therefore because it serves as a means by which system sustainability can be enhanced. It does this through facilitating decision-making, particularly on the quality of service and cost recovery policy which reflects what people want and are willing to pay.

SSA advocates willingness to pay surveys (WTP) as the main method to assess demand. Alternative approaches, such as affordability rule of thumb and benefit transfer have been found to lead to serious inaccuracies (DFID, 1997). For example, an affordability approach neglects research indicating that income is only one of several determinants of willingness to pay (Whittington et al, 1992). Differences in characteristics of supply, including cost, reliability and quality, between the improved and alternative sources of supply are also important, as are the socio-economic characteristics of the household and attitudes to government policy (World Bank Water Demand Research Team, 1993). In general, although these alternative approaches may give an indication of the current costs consumers bear for services, they fail to take account of the benefits from improved services, or how many households will use the improved system (Department for International Development (DFID), 1997).

The contingent valuation method (CVM), and its variants, have received much attention in the literature on demand assessment. In particular, Altaf and Hughes (1994) and Whittington (1998) have reviewed the use of CVM in measuring demand for services (in most cases relating to water supply) in developing countries. Several key advantages over other techniques can be identified; it can be used to study willingness to pay at different levels of service not currently available, it can help to determine the full range of benefits from service improvements, and can indicate how many households will switch to a new system.

But the constraints on the use of the technique are significant. Specialist expertise in the framing of CVM questionnaires, in result analysis and interpretation is generally required, thereby increasing its cost (a thorough CVM study is estimated to cost at least £100,000), and reduces the potential for community based involvement in its administration. There are also question marks about the reliability of the method, particularly relating to the hypothetical nature of questions asked, and the danger that information obtained through community meetings may be subject to 'mob' effects, leading to bias. Cotton (1998) argues that the emphasis on demand assessment fails to recognise the very different circumstances in which these techniques have been applied. CVM and pure demand based approaches work well in the context of water supply schemes, deciding on standposts, yard connections and house connections, but these same techniques do not apply with the same clarity to sanitation projects. For example, in cases where only a few households receive a latrine on a demand led basis, the benefits of such service improvement are minimal if the households' neighbours are defecating in the open and contaminating the local environment. There are also unresolved issues relating to how demand is negotiated after it is expressed; what if some members of the community choose sewerage and others latrine based systems? How should the technical considerations inherent in these choices be integrated into community based expression of demand?

In order for demand assessment studies to realise their full potential (i.e., to inform policy choices at the strategic framework level, and to help shape the design of individual projects) it is essential that they are used at more than one stage in the project development process, and that they are integrated in such a way that their findings influence design choices (DFID, 1997). It is also observed that demand driven planning processes pose considerable implications for consultants, planners, economists and government officials, in that it requires them to be more flexible in their responses to user communities (Whittington et al, 1996). Consultants need to be willing to adopt an open minded approach to problems of basic service delivery, and consider different service scenarios for a community, based in part on the community's own desires. This is difficult in that there are few incentives for consultants to work in this way.

A central weakness of the emphasis placed on the need for demand driven approaches is the significant institutional change they imply, and the ability of those institutions to respond to a changed institutional culture. Demand driven approaches involve different processes than the formal supply driven process of urbanisation (WASH, 1993). Typically, urban sector institutions in developing countries are set up for the latter, and are not organisationally structured and prepared to carry out demand driven urbanisation. Institutions have their own laws and regulations, institution mission statements, goals, personnel and methods. Staff are trained in a supply oriented manner in their approach to urban planning, thus discussions of demand driven approaches often lead to confusion for staff about institutional and personal responsibilities towards peri-urban communities. An additional problem with the approach is that assessing demand becomes an exercise which can inflate expectations about service delivery, which could be years in delivering, or simply not feasible.

A key element missing from SSA is recognition that demand for sanitation is more than a one-dimensional concept. RWSG - South Asia (1997) describe two levels of the concept, beginning with unexpressed or latent demand. Although some projects, such as the Bangladesh Slum Improvement Project (SIP) may appear to be essentially supply driven, the fact that there was no formal assessment of sanitation demand does not imply that there was no demand for sanitation services. Demand may be present but remain unexpressed or latent. A key question then becomes how this latent demand can become explicit. A second division focuses on uninformed and informed demand. In the Orangi Pilot Project, Pakistan there was a clear expression of community demand for a solution to sanitation and drainage problems, but little consideration of concern for the wider environment. Discharge from Orangi’s sewers flowed into open nullahs and polluted the Lyari river. Community demand, although explicit, was partially informed in that it failed to appreciate the impact on the downstream environment. Similarly, demand may be uninformed in cases where there is no consideration of the true financial and economic costs of an option, or possible technical difficulties associated with implementing particular sanitation options.

A related point is that demand is often based on what people know or see around them. How then is it possible for the community to demand options which they are unfamiliar with? Demand for a product, good or service does not occur in a vacuum; it is affected by the local context. Thus, RWSG-SA argue that there can never be a purely demand driven approach in the choice of technology because this must be influenced by the availability and cost, which are both supply side factors. For example, in Orangi, the choice of sewers was affected by the topography of the area which allowed gravity discharge to nullahs.

Key points:

Key questions:

Unbundling sanitation investments
'Unbundling' as a concept is a way of dividing investments into more realistic and manageable components, allowing for more appropriate and efficient use of resources.

SSA describes two types of unbundling. Horizontal unbundling involves the subdivision of services/technologies geographically, so that a city may be broken down into two or more zones, each with self contained and independent sanitation services. Dividing sanitation systems into zones helps to reduce the average diameters and depths of sewers when compared with a single centralised system, and as these are likely to be the two highest cost elements, it is probable that horizontal unbundling is sound economically wherever it is technically feasible. Unbundled systems may also be linked, for example at a treatment works or long outfall. An example of horizontal unbundling quoted by Wright is the case study of Manila, the Philippines, where the metropolitan area has been horizontally unbundled into two parts for the supply of water and sewerage services.

Secondly, there is vertical unbundling, where sanitation programmes are divided according to the scale and cost of the components. Division may apply to the design and construction phases, and to the O&M of the systems. Further subdividing is possible to permit community managed schemes to connect into publicly or privately operated sewerage systems. The key advantages of vertical unbundling are that it may provide flexibility in the financing of urban sanitation, and help to make connections affordable through equitable cross subsidy. It is particularly valuable in that it helps to separate investment decisions on in-house, feeder and trunk levels thereby allowing a clear link to be made between benefits and costs. The condominial sewerage systems of Brazil are cited as an example of this form of unbundling.

Unbundling as a whole may lead to progress on the development of city wide sanitation programmes that would have stalled if financed and implemented as a single package (Wright, 1997).

In sanitation programmes, unbundling is a new term which to an extent describes established practices. Horizontal unbundling typically takes place when different technological options are used in separate parts of a city, such as when on-plot systems are used in some districts, and conventional water borne sewerage in others. Likewise, vertical unbundling can be found in many sanitation systems in that services within the plot boundary are normally the householders' responsibility, while those beyond the plot are in the public domain and provided by local authorities or government.

In the proceedings for a regional workshop on sanitation for low income urban communities, RWSG - South Asia (1997) make two general observations about the unbundling concept. The first reinforces the idea that unbundling is not a new concept; what is new is the manner in which it is described. Clear similarities exist between unbundling and the idea of 'moving the boundary' between private/community sector provision. It has become accepted practice in development that it is best to leave responsibility for facilities and services within the plot to householders, with a support service role being fulfilled by the public sector. 'Moving the boundary' recognises that local stakeholders are well placed to undertake the roles of provision and management of some services beyond this traditional plot boundary division. In consequence, the public sector's role is shifted towards the connections between secondary and tertiary facilities. Parallels between this shifting of responsibilities and the idea of disaggregating roles through an unbundled approach are clear.

Secondly, RWSG - South Asia observe that with the unbundling of responsibilities there is greater scope for independent activity at the local level than is generally the case with higher order facilities. This is illustrated through the example of a hypothetical sewerage scheme: provision of sewers at the local level will be dependent on the presence of suitable secondary sewers for receipt of discharge. Hence the need for an integrated approach. However, tertiary sewers can be provided independently in streets and lanes once secondary sewers and appropriate disposal facilities have been constructed. The critical point is that the closer to the household level, the greater the scope for horizontal unbundling to take place.

Horizontal unbundling is possible when part of a sanitation system is managed independently of other facilities at the same level. RWSG - South Asia (1997) cite two examples of projects in which such unbundling has occurred: Sukkur, India, and Sri Lanka. In each case, faecal waste and sullage from each district are handled separately from sanitation networks in adjacent urban areas. For example, in the Sri Lankan example, sewers discharge to the settlement-level septic tanks. Additionally, the OPP experience in Pakistan reinforces the opportunities for horizontal unbundling at the lower or local end of vertically integrated systems, despite centralised control over higher order facilities. In this project, horizontal unbundling was observed at lane level, with individual lane organisations drawing up different strategies to manage the sewerage system (Siddiqui and Rashid, 1997).

There is little discussion in the literature of appropriate conditions for unbundling. Cotton (1998) however, suggests that certain variables can limit its opportunities. In Cuttack, India, the existing sewerage system is too small to consider unbundling; the topography of Cuttack and the consequent need for pumping makes it difficult to see how locally decentralised sanitation systems would be viable. The lesson is that in some circumstances, cities may not be suited to unbundling.

Vertical unbundling is to some extent inherent in all sanitation systems in which a hierarchical system exists in which facilities and services are provided at a number of levels. This hierarchy is most explicity seen in a sewered system where on-plot facilities are connected to branch sewers which are connected to collector and trunk sewers which direct sewage to some form of treatment before being discharged into the wider environment. Wright argues for a three level distinction of facilities, in-house, feeder and trunk levels. OPP differentiate a four level typology including in-house (on-plot), local feeder, secondary feeder/collector and primary/city wide facilities. The latter definition has the advantage of distinguishing between tertiary level facilities that would appear to present opportunities for community action and secondary level facilities which would appear to be best provided by central authorities.

The Bangladesh SIP illustrates vertical unbundling on two levels. Initially, the project separated responsibilities between a project management unit (responsible for services within the project) and government agencies which continued to deliver services to the city as a whole. Furthermore, within the project management unit there existed a four tier structure (including women's groups, sub-project implementation committee, project implementation committee and central co-ordination committee), which attempted to further unbundle responsibilities between these particular sections.

Further lessons from vertical unbundling can be found from OPP, where it was found that the effectiveness of devolving responsibility for finance, management and maintenance of the sanitation system depends critically on institutionalising community mobilisation and training procedures and on the degree of transparency in working with government agencies (Siddiqui and Rashid, 1997).

The collection of excreta from large numbers of on-plot sanitation systems is a sanitation activity which can be organised in such a way as to allow for unbundling of responsibilities (Muller, 1997). In an analysis of four case studies from Ghana, Tanzania, India and China, a wide variety of actors were found to be employed in the collection process, including informal, small scale enterprises, local scavengers, formal private enterprises and even municipal labourers. The process involves four stages: removal/primary transfer to neighbourhood collection point; secondary transfer; treatment and disposal. In general, the role of private formal/informal organisation tends to concentrate on the removal/primary transfer stage of the process, with secondary transfer, treatment and disposal predominantly the responsibility of municipal departments, although in Tanzania, large scale informal waste collectors were involved in treatment. As with other experiences of unbundled systems, Muller stresses the need for linkages between the separate parts, arguing that, "...primary collection must be linked to a larger sanitation system that operates facilities for treatment and disposal of excreta " (Muller, 1997:71). The internal functioning of municipal sanitation departments, their capacity to manage personnel, equipment and finance is viewed as a key element in influencing its ability to interact with other organisations in the process.

The potential benefit of unbundling responsibilities is demonstrated in a case study of innovative approaches to solid waste management in three Bangladeshi municipalities (Gaffar and Rahman, 1997). In this case, the solid waste chain was unbundled according to physical handling responsibilities, and the key outcomes recorded were a 20-25 per cent reduction in budget costs during the two year operation of the project, an improvement in service provision for householders (drains were cleared, waste was removed from streets), positive reactions to the project from users.

Although these case studies demonstrate the way in which either horizontal or vertical unbundling has been achieved, the literature review provides several examples of risks associated with unbundling. Matin (1997) when commenting on the Bangladesh SIP argues that vertical unbundling had mixed results in this project, primarily because the management structure meant planning decisions tended to be dominated at the programme implementation committee and central co-ordination committee level. Community level groups, such as women's groups and sub-project implementation committee tended to implement what had been decided at higher levels.

One of the weaknesses from unbundling is that little attention is paid to how links are made between disaggregated systems. Cotton (1998b) cites experiences from the Department for International Development (DFID) sponsored Cuttack Urban Services Improvement Project (CUSIP) which illustrate this point. The three key agencies with a stake in urban sanitation in Cuttack include the CUSIP, Cuttack Municipal Corporation (CkMC) and the Public Health Engineering Department (PHED). Each agency has adopted different programmes for sanitation, and these are unbundled to the extent that there are different institutional responsibilities and financing arrangements for on-plot, neighbourhood and centralised infrastructure. The key problem for Cuttack, and a lesson for SSA as a whole, is that unbundling can only take place in the context of a strategic framework, where there is some provision for co-ordination of activities. Rather than arguing for further unbundling, there is a need for 'bundling up' in some cases, to ensure that the each agency is aware of each others' roles and responsibilities. An additional concern raised from this point is that of the institutional capacity required to deal with unbundling, and resolution of questions such as who oversees the process and who co-ordinates the constituent parts.

Tayler (1997) cites experiences of sewerage systems in Pakistan that have been unbundled with responsibilities shared between agencies, leading to sustainability problems with the systems. In Northeast Lahore, the design of tertiary and some secondary sewers was the responsibility of the upgrading unit in the Metropolitan Planning Wing of the Lahore Development Authority, whilst ongoing O&M was assigned to Lahore Water and Sanitation Agency (WASA). Two problems arose from this division of responsibility: WASA's failure to complete construction of off-site sewers and a reluctance to accept 'appropriate' design standards advocated by the designers. Hasan (1997) reports a similar problem in Sukkur, where a PHED designed main trunk sewer had been inoperable for 15 years, due to partial sewer collapse and the failure of pumping stations. Although PHED received funds for new works, no provision was made for repair and O&M from Sukkur Municipal Corporation. This may be because no formal arrangement was made between key stakeholders regarding financing of the project running costs. Hasan comments that collaboration between SMC-PHED could have been achieved when the funding allocations were being made by the provincial government. However, no such co-ordination was present.

Both the Bangladesh SIP and the OPP projects have experienced problems with linking aspects of their unbundled systems. In Bangladesh, SIP drains and the city wide drainage systems do not coincide, leaving SIP drains discharging pools of polluted water on the periphery of the slums (RWSG - South Asia (1997)). Similarly, there has been little progress to date on provision of trunk sewerage in Orangi.

Key points:

Financing and cost recovery
SSA argues for a central role to be given to user charges with regard to financing. This means that users are encouraged to contribute according to their willingness and ability to pay for the services they have chosen as best meeting their needs. The rationale for this approach is based on the lessons learnt from previous city wide, donor financed projects. Typically, these schemes were dependent on widespread subsidies to get projects off the ground because it was assumed that users did not have adequate means to pay. The consequence was unsustainable interventions.

The demand based approach to SSA provides a way to think through how the costs of sanitation can be best shared, and it also provides reliable information on the value that users place on sanitation improvements and on their WTP for those benefits. Wright advocates a model of financing sanitation improvement which is based on appropriate cost recovery at household, block, neighbourhood, city and river basin level.

Direct user payments: User WTP is generally limited to those benefits that users perceive and are able to internalise. User WTP may be insufficient to pay for the full cost of systems, including trunk sewers and treatment, thus, some form of complementary finance will be required. The examples of OPP and PROSANEAR show that user WTP should be tested before considering cross subsidies from other user groups or external provision of funds. Not all direct user payments need to be in cash, but in-kind contributions have to be voluntary and must tangibly reduce the real financial cost of providing the service. Household borrowing to finance user payment is acceptable - the most effective system for ensuring user payment for sanitation have been those insisting on front end payments for construction and access to the service. Examples of credit systems are known, including Lesotho’s low cost urban sanitation scheme.

Collective payments from block and neighbourhoods: Neighbourhood groups may be useful in organising collective payments (although there may be problems in group decision making). In OPP, neighbourhood groups helped provide finance for neighbourhood collection systems. In both OPP and PROSANEAR, groups were able to encourage individuals to sign on to less costly neighbourhood schemes, and to undertake the O&M for them, than the existing sanitation organisations had been able to do. Neighbourhood groups were instrumental in convincing technical staff that cheaper options could work well, at costs that households could afford.

Collective payments from local and national governments: Directly or indirectly government participates in the financing of sanitation services in all countries (i.e. financial support to users, allocation of funds for investments, taxes and control of credit allocations).

The financial premise of SSA is to base financial requirements on what is worthwhile to finance, rather than assuming that if something can be financed, it is worthwhile. Government finance has been extensively used in the past as a substitute for user charges and has benefited those who already have access to services, rather than compensating service providers for external benefits of extending services to users who are not willing or able to pay. One outcome of this is a continuing reluctance to extend service coverage to the poor; also relatively little expansion of treatment facilities to protect water sources and insufficient maintenance of existing facilities for those with connections (Wright, 1997).

The shift in development thinking away from supply-led financing strategies to those that are demand-based implies that if the financial element of a sanitation programme is to be sustainable, then considerable information about the financial context in which communities operate will be required. This includes, "...information relating to the availability of credit facilities, the willingness of the communities to pay for sanitation, government attitudes towards cost recovery, the role of the private sector and so on ". (Cotton and Saywell, 1998:35).

SSA's premise on financing and cost recovery is that sound finances are best achieved where costs are borne either directly or indirectly by the users of sanitation services.

OPP illustrates one aspect of this point, relating to capital costs, with the users of tertiary sewers in Orangi managing their construction and paying their full cost, without subsidy or loan arrangement. Construction costs for lane level and secondary sewage lines in Orangi were approximately Rs. 1000, collected through lane level organisations charged with the responsibility. This method of cost recovery has an additional benefit of offering transparency / accountability of financing arrangements at the local level. (Siddiqui and Rashid, 1997). A variation on the Orangi approach is that adopted in the Faisalabad Area Upgrading Project (FAUP), where tertiary level sewers were paid through equal contributions from users and government inputs to the FAUP (UNDP-World Bank Water and Sanitation Group RWSG - South Asia (1996)).

Many sanitation projects attempt to recover costs from users. However, approaches vary considerably, and the effectiveness of recovery schemes may be adversely affected by the local operational context. NSDF/SPARC/MM found in Dharavi, Bombay that it was difficult to implement effective cost recovery schemes in situations where users had become used to receiving services free of charge (UNDP-World Bank Water and Sanitation Group RWSG - South Asia, 1997). Additionally, there are problems associated with cost recovery schemes which operate alongside projects where no attempt at cost recovery is made.

SSA argues that borrowing to finance direct user payment is acceptable, although relatively little attention is paid to the use of credit mechanisms, beyond citing the Lesotho sanitation scheme. Cost recovery in sanitation projects is achievable, and credit schemes offer a system for attaining this goal. Varley (1997) cites several lessons learnt from household credit financing schemes for sanitation, including:

Additionally, Cotton and Saywell (1998) abstract key lessons learnt from a variety of cost recovery schemes relating to the provision of household level sanitation facilities:

Although OPP and FAUP demonstrate that direct cost recovery of capital costs through user contributions is possible, there are still questions relating to community preparedness to pay for on-going running costs of the systems. In OPP it was evident that many users were unwilling to assume the necessary responsibility for financing recurrent costs of maintenance given the frequency with which the systems had to be cleaned and the rising charges made by scavengers for this maintenance service. In consequence, it was found that regular cleaning of manholes, pipes and replacement of covers was often left unattended.

On a more general level, the debate surrounding the financing of urban services is addressed specifically by Rondinelli (1990). The transfer of administrative and financial responsibilities for urban services to municipalities is seen as one way in which increasing demands on services can be met, at the same time as relieving some of the fiscal pressure on central treasuries. However, decentralisation alone is not a panacea for financial problems. In particular, central government will be unable to decentralise without first strengthening municipal administrative capacity, and addressing the problems of poor organisational structures and operating procedures.

Cotton (1998b) identifies a conflict in the theoretical assumptions behind the financing and cost recovery elements of SSA. On the one hand, emphasis is given to a 'pay as you use' approach, where in order to be equitable all costs (capital, recurrent etc.) need to be contained in the user charge (connection fee and recurrent payments). However, when describing the provision of trunk infrastructure, SSA suggests that general taxation should cover not only capital but also O&M costs. This has the effect of encouraging a hidden subsidy.

Key points:

Key questions:

Miscellaneous issues

Small steps approach
The small steps approach in relation to SSA implies a recognition of the links between physical infrastructure provision and the existing institutional/financial capacity to cope with such activities. It also allows for lessons learnt from experience to be reinvested into the on-going planning process. Such an approach is seen as fundamental if sanitation provision is to be shared process of learning rather than the application of discrete solutions.

Relatively few sources in the literature explicitly discuss this matter. UNDP-World Bank Water and Sanitation Group RWSG - South Asia (1997) is one exception to this rule, citing case studies where such an approach has been tried:

Key points:

Key questions:

Strategies of institutional design: broadening competition
One of the assumptions underpinning SSA's is that broadening competition is a mechanism by which operational efficiency can be achieved with minimal transaction costs.

Recent research findings question this assumption. Batley (1996) investigated the extent to which private sector involvement led to improved delivery of a range of urban services (although admittedly not sanitation) in six developing countries. The study made comparisons of the ways in which local level services were provided in cities in Brazil, Mexico, Uganda, Zimbabwe, India and Malaysia. Although in some cases there was evidence to suggest better private performance in service delivery, it was not possible to conclude that privatisation automatically led to greater efficiency and effectiveness overall. Key points arising from the research with relevance to SSA include:

In practice it may be difficult for weak public administration to undertake new roles which follow from partial privatisations, such as setting policy frameworks, managing contracts, regulating contractors and financing/supporting producers and consumers.

Updated 31/01/03

Maintained by f.o.odhiambo@lboro.ac.uk and j.fisher1@lboro.ac.uk